X-Git-Url: http://git.rot13.org/?p=MARC-Fast;a=blobdiff_plain;f=Fast.pm;h=b587be3d0db5042e96239a28f2d2de944434988d;hp=6116b1494450e2701f8b37fa48418998af59c3a7;hb=12f18ad14a52dc83e3449d908fb0678e16dabb49;hpb=8300e4b53e83d8a355dba26d74624d521f67232f diff --git a/Fast.pm b/Fast.pm index 6116b14..b587be3 100644 --- a/Fast.pm +++ b/Fast.pm @@ -1,5 +1,5 @@ - package MARC::Fast; + use strict; use Carp; use Data::Dumper; @@ -7,7 +7,7 @@ use Data::Dumper; BEGIN { use Exporter (); use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); - $VERSION = 0.02; + $VERSION = 0.08; @ISA = qw (Exporter); #Give a hoot don't pollute, do not export more than needed by default @EXPORT = qw (); @@ -23,12 +23,21 @@ MARC::Fast - Very fast implementation of MARC database reader use MARC::Fast; + my $marc = new MARC::Fast( + marcdb => 'unimarc.iso', + ); + + foreach my $mfn ( 1 .. $marc->count ) { + print $marc->to_ascii( $mfn ); + } + +For longer example with command line options look at L =head1 DESCRIPTION This is very fast alternative to C and C modules. -It's is also very sutable for random access to MARC records (as opposed to +It's is also very subtable for random access to MARC records (as opposed to sequential one). =head1 METHODS @@ -42,6 +51,11 @@ Read MARC database quiet => 0, debug => 0, assert => 0, + hash_filter => sub { + my ($t, $record_number) = @_; + $t =~ s/foo/bar/; + return $t; + }, ); =cut @@ -106,7 +120,7 @@ sub new { print STDERR "REC ",$self->{count},": $leader\n" if ($self->{debug}); # store leader for later - push @{$self->{leaders}}, $leader; + push @{$self->{leader}}, $leader; # skip to next record my $o = substr($leader,0,5); @@ -140,14 +154,22 @@ Fetch record from database my $hash = $marc->fetch(42); +First record number is C<1> + =cut sub fetch { my $self = shift; - my $rec_nr = shift || return; + my $rec_nr = shift; - my $leader = $self->{leaders}->[$rec_nr - 1]; + if ( ! $rec_nr ) { + $self->{last_leader} = undef; + return; + } + + my $leader = $self->{leader}->[$rec_nr - 1]; + $self->{last_leader} = $leader; unless ($leader) { carp "can't find record $rec_nr"; return; @@ -229,6 +251,26 @@ sub fetch { } +=head2 last_leader + +Returns leader of last record Led + + print $marc->last_leader; + +Added in version 0.08 of this module, so if you need it use: + + use MARC::Fast 0.08; + +to be sure that it's supported. + +=cut + +sub last_leader { + my $self = shift; + return $self->{last_leader}; +} + + =head2 to_hash Read record with specified MFN and convert it to hash @@ -265,14 +307,14 @@ sub to_hash { my $row = $self->fetch($mfn) || return; - foreach my $k (keys %{$row}) { - foreach my $l (@{$row->{$k}}) { + foreach my $rec_nr (keys %{$row}) { + foreach my $l (@{$row->{$rec_nr}}) { # remove end marker $l =~ s/\x1E$//; # filter output - $l = $self->{'hash_filter'}->($l) if ($self->{'hash_filter'}); + $l = $self->{'hash_filter'}->($l, $rec_nr) if ($self->{'hash_filter'}); my $val; @@ -283,30 +325,57 @@ sub to_hash { if ($l =~ m/\x1F/) { foreach my $t (split(/\x1F/,$l)) { next if (! $t); + my $f = substr($t,0,1); + # repeatable subfileds. When we hit first one, + # store CURRENT (up to that) in first repetition + # of this record. Then, new record with same + # identifiers will be created. + if ($val->{$f}) { + push @{$rec->{$rec_nr}}, $val; + $val = { + i1 => $val->{i1}, + i2 => $val->{i2}, + }; + } $val->{substr($t,0,1)} = substr($t,1); } } else { $val = $l; } - push @{$rec->{$k}}, $val; + push @{$rec->{$rec_nr}}, $val; } } return $rec; } +=head2 to_ascii -1; -__END__ + print $marc->to_ascii( 42 ); -=head1 BUGS +=cut + +sub to_ascii { + my $self = shift; + my $mfn = shift || confess "need mfn"; + my $row = $self->fetch($mfn) || return; + my $out; -=head1 SUPPORT + foreach my $f (sort keys %{$row}) { + my $dump = join('', @{ $row->{$f} }); + $dump =~ s/\x1e$//; + $dump =~ s/\x1f/\$/g; + $out .= "$f\t$dump\n"; + } + return $out; +} +1; +__END__ =head1 AUTHOR @@ -326,6 +395,6 @@ LICENSE file included with this module. =head1 SEE ALSO -perl(1). +L, perl(1). =cut